Storm Surge Issues of Hurricane Katrina

1
Storm Surge Issues of Hurricane Katrina P.
Fitzpatrick1, Y. Lau1, S. Bhate1, Yongzuo Li1,
Elizabeth Valenti2, Bob Jacobsen3, and Joel
Lawhead4 1. GeoResources Institute, Mississippi
State University 2. WorldWinds Inc. 3. URS 4
NVision Solutions Inc.
Timing of Wind and Surge An important insurance
issue involves the timing of wind versus surge.
All tide gauges failed at the peak of the storm
in the severely impacted regions. However, 17
USGS gauges in the impact region functioned
during tropical storm-force conditions. Figure 3
shows a typical gauge measurement at Bay Gardene,
LA. All gauges show winds of 50-60 mph with
storm surge values of 5-8 feet, typically less
than would flood most homes. Printouts of all
tide gauges are available for examination at this
poster station.
Overview Category 3 Katrina generated a
U.S.-record storm surge which impacted a wide
region from Grand Isle, LA, to Mobile Bay, AL,
and killed about 1350 people with hundreds still
missing. Sensitivity experiments by WorldWinds
using the ADvanced CIRCulation (ADCIRC) storm
surge model show a large hurricane produces water
elevations 20-40 higher than a small hurricane
with the same intensity and with considerably
more widespread inundation. The previous
U.S.-record surge was Hurricane Camille (1969)
which impacted the same region, an intense but
smaller Category 5 hurricane. Camilles
hurricane-force winds extended 60 miles from the
storm center, while Katrinas extended 120 miles.
Camilles tropical storm-force winds reached 180
miles outward, while Katrina was 230 miles. It is
likely Katrinas wide eye of 37 miles also played
a role (Camilles eye was approximately 11 miles
wide). The inland penetration of Katrinas storm
surge was truly remarkable. The Mississippi River
levee system held and confined most of the surge
east of the river except for the landfall region
of Buras, LA. Regions west of the Mississippi
River experienced little surge, suggesting that
the river levee system may have augmented
Katrinas surge on the east side. Most of
Plaquemines, St. Bernard, and eastern Orleans
Parishes were inundated with surge which
overflowed levees and destroyed them with
scouring action. Buildings outside the levee
system became cement slabs. Tide gauges also show
the surge traveled up the Mississippi River, with
elevation spikes reaching 14 feet at the Bonnet
Carre Spillway 10 miles west of New Orleans.
Levees along some canals south of Lake
Pontchartrain were not overtopped but experienced
failures that are still under investigation,
causing well-publicized flooding of New Orleans.
The surge also penetrated through inoperative
flood pumps which, when combined with the
inability to remove rainwater, caused moderate
flooding in the suburban region west of New
Orleans. The eastern end of St. Tammany Parish
suffered an extreme surge which came from Lake
Borgne as well as up the Pearl and Bonfouca river
systems, traveling miles inland in Slidell. St.
Tammany experienced a second surge when the wind
shifted, sloshing piled-up water in Lake
Pontchartrain northeastward. A video is shown of
the surge along northern Lake Pontchartrain.
Table 1. Hurricane Katrina high water marks,
compiled by Haag Engineering, National Weather
Service (NWS), the USGS, and Fitzpatrick along
the immediate coast. Available Hurricane Camille
values from the Army Corps of Engineers is also
shown for comparison.
145
125
105
85
65
45
25

Fig. 3. USGS tide gauge of water elevation (feet)
and wind (mph) for Bay Gardene, located in the
marsh 20 miles east of Chalmette, LA. This gauge
failed after midnight on 8/29/05.
Fig. 4. ADCIRC simulation of the surge (feet) and
the wind forcing (mph) for an inland point near
Bay St. Louis, MS during landfall.
To examine the possible timing of surge in the
storm center, time series plots of individual
locations were produced from the ADCIRC
simulation. They generally show the peak winds
preceding the peak surge between 30 minutes and
two hours. An example is shown for Bay St. Louis
(Fig. 4).
The Impact of the Mississippi River Gulf Outlet
and the Hypothesized Funneling Effect The
Mississippi River Gulf Outlet (MRGO) is a
70-mile, deep-draft, man-made channel, completed
in 1963. 40 miles was dredged through marshland
in St. Bernard Parish. Originally 750 feet wide,
MRGO has eroded to 2000 feet wide in many places,
destroyed more than 36,000 acres of wetlands, and
disrupted a brackish environment with high
salinity. Its role in hurricane storm surges is
also controversial, with speculation MRGO acts as
a conduit. Its intersection with the Gulf
Intracoastal Waterway has also been hypothesized
to provide a funneling effect for the storm
surge. These issues were studied with an ADCIRC
simulation of Katrina for a filled-in MRGO.
Little difference was observed (Fig. 5), because
the surge is a widespread event in which one
channel will make little impact. Another
simulation used a wider levee system at the
intersection, and the surge actually increased
(Fig. 6), because without restricting the flow,
the conveyance increased in this region. The URS
report is available at this poster station.
Katrina storm surge at 5AM
Katrina storm surge at 7AM
The entire Mississippi coast experienced the
storm surge. The western region from Pearl River
to Bay St. Louis suffered the worst, as the surge
traveled past Interstate 10 (Figure 1). The
official peak surge occurred in this region,
estimated at 28 feet. However, high water marks
indicate even higher elevations, although some
may be impacted by wave action (Table 1).
Comparisons to Hurricane Camilles surge are
shown in Table 1. The surge also traveled far up
the Jordan River and Biloxi River, decimating
towns such as Kiln, MS. An ADCIRC simulation of
Katrinas surge evolution is shown in Fig. 2. The
surge occurred at high tide, adding another foot
of water.
Stennis Space Center
Katrina storm surge at 9AM
Katrina storm surge at 11AM
Jordan River
I-10
Waveland
Pearl River
Fig. 2. ADCIRC simulation of Katrina storm surge.
Graphics extend from the Northshore of Lake
Pontchartrain, LA, across the Mississippi coast
to Pascagoula, MS. At 5AM the surge moves up the
Pearl, Jordan, and Biloxi River estuaries. Marsh
regions near Pearlington and Pascagoula begin to
experience inundation. The surge is below 5 feet
in most regions. By 7AM, this pattern continues,
but with surge values above 10 feet in some
regions. Waveland begins to experience
inundation. By 9AM, significant storm surge is
occurring along the Mississippi coast and Mobile
Bay, with 15-25 feet water elevations penetrating
miles inland west of Bay St. Louis. Because the
wind direction is shifting over Louisiana,
piled-up water in Lake Pontchartrain pushes
eastward, causing a second wave of inundation in
that region. Indeed, damage to the twin spans
bridge system which connects Slidell and New
Orleans, indicates an outward surge, with much of
the damage on the east of the bridge system. The
peak surge occurs around 11AM period, with
extreme inland penetration and record surge
values on the order of 25-35 feet. Even though
Katrina was less intense than Camille, the record
surge can generally be explained by the huge size
of the storm.
Fig. 1. The maximum storm surge in Hancock County
according to FEMA. Some high water marks also
shown. The shaded region represents the inland
penetration. Note the surge traveled past
Interstate 10 and up the Jourdan River and Pearl
River. (Figure courtesy of Nvision Solutions
Inc.)
Fig. 5. Comparison of ADCIRC simulations with
MRGO open and closed, showing little impact of
the MRGO on the storm surge at Bayou Dupre south
of Lake Borgne near Violet, LA.
Fig. 6. Comparison of ADCIRC simulations with
MRGO closed and a wider levee system at the Gulf
Intracoastal Waterway. A wider levee system
actually yields higher surge elevations, because
the conveyance increases.